Re: [AMBER] How should I perform a long-time Simulated Annealing MD?

From: David A Case <>
Date: Wed, 4 Jul 2018 22:06:11 -0400

On Wed, Jul 04, 2018, Robert Molt wrote:

> I was hoping to ask a follow up question to one specific element of Dr.
> Case's answer. Dr. Case wrote:
> "The time scale for many such changes is on the microsecond to
> millisecond time scale (or longer), and are hence not readily approached
> by brute force simulations. (And, of course, current force fields may
> not properly describe all of these alternate conformational basins.)"
> I am confused by the parenthetical statement. Is there a particular
> reason to believe that the quality of a force field varies by a
> conformational basin? I would assume if one starts in basin A, the force
> field should not be assumed to be more accurate there than in basin B?
> Or is this just meant as a general "snipe" of force field limitations?

I think it's a genuine problem: the relative (free) energies of different
conformational basins may not correspond all that well to experiment. If
one begins in state B and expects to see a spontaneous transition to state
A, one is hoping that state A is indeed favored with the force field being
used. Since allosteric states are typically close in energy, it doesn't
take much of an error to reverse their relative energies from what they
"should" be.

(To be yet more parenthetical, :-), you can swap labels "A" and "B" in the
above argument and reach much the same conclusion. cf. W.S. Gilbert, See how
the Fates their gifts allot.)


AMBER mailing list
Received on Wed Jul 04 2018 - 19:30:02 PDT
Custom Search